Difference Analysis On The Online Language Features Between Higher And Lower Level Of Internet Dependency Users

With the rapid development of the Internet, the phenomenon of excessive Internet use has been widely concerned by psychologists. Many researchers have studied Internet addicted behavior from different perspective, such as cognitive and emotional dysfunction. It is generally accepted that Internet addicted users is characterized as poor performances in their daily lives. Previous studies analyzed and identified excessive Internet use mainly through the observation of users’ offline behavior; however, the empirical research concerning Internet addicts’ online behavior is insufficient. In this study, we explored the online language features between the higher and lower level of Internet dependency users by inducing the individual’s emotional state,selecting different kinds of internet cues and combining the text analysis techniques, and aimed to provide a value reference for the future online screening of Internet addiction.In study 1, we compared the differences of online language words between the higher and lower level of Internet dependency users by inducing the positive and negative emotional state and reading the positive and negative news texts. The results showed that individuals’ emotional state has little impact on the online language usage of users with higher level cyber-dependency, but the emotional attributes of the reading text have a significant effect on it. Meanwhile,the individual’s negative mood experience can be improved through cyber activities. Especially for users with higher level cyber-dependency, engaging in internet activities make them easily get more positive self-experience.In study 2, through simulating the common graphic-reading mode in social media,we explored the online language features with different degrees of Internet dependency users in a more realistic network situation. The results showed that, graphics can affect the online language usage of users with higher level cyber-dependency, especially cartoon emoticons which is highly connected with the Internet. For users with higher level cyber-dependency, they are more familiar with cartoon emoticons; if the type of corresponding pictures is cartoon emoticons, their re-evaluation of positive text information will be hindered but the re-assessment of the negative text information will be aroused. Furthermore, corresponding cartoon emoticons allowed higher level Internet dependency users reduce the negative feelings of text information. However, for users with lower level cyber-dependency, their adverse emotional experience about negative text information became more serious when they read news texts with cartoon emoticons.In summary, for users with higher level cyber-dependency, their online behavior are significantly different to offline behavior. More advanced data mining technology should be utilized in the future researches to dynamically identify the Internet addiction behavior by first determining the reading texts’ emotional attributes and their corresponding graphs types, and then exploring the language features of topic-based comment text.